Today’s National Statistics decision on the retail prices index (RPI) has provoked outrage from statistical purists but represents a sensible, if untidy, compromise.

Against forecasts, the National Statistician has addressed criticism of the calculation method used for the RPI by introducing a new measure, RPIJ, that employs superior methodology, rather than changing the RPI itself. The creation of an additional inflation measure risks causing confusion but avoids the windfall gains / losses entailed by reengineering the RPI. (The change would have lowered RPI inflation, benefiting the government and some pension scheme sponsors at the expense of holders of existing index-linked debt and recipients of RPI-linked pensions.)

The economic importance of the RPI has waned in recent years because of government decisions to switch to the CPI for uprating benefits and taxes. The introduction of RPIJ will continue this trend, as will the inclusion of owner occupiers’ housing costs in a new CPI measure, CPIH, to be released in March. CPIH may assume eventual dominance, having similar broad coverage to RPIJ but a superior weighting structure, reflecting the spending patterns of all UK households plus foreign visitors, whereas the RPI / RPIJ excludes the highest earners and some pensioners.

The Treasury has announced that it will continue to issue new index-linked gilts based on the RPI. There would be no benefit from switching to RPIJ / CPI / CPIH since investors would demand a higher initial real yield to compensate for lower future inflation compensation payments.